Automated natural language communication analysis

Abstract

A device may receive information identifying a communication framework for a mass communication task. The device may determine a success score for the communication framework using a mass communication model, wherein the success score represents a likelihood of a successful response in connection with using the communication framework for the mass communication task. The device may generate a recommendation for the communication framework based on the success score and using the mass communication model. The device may alter the communication framework to implement the recommendation and generate a modified communication framework. The device may perform the mass communication task using the modified communication framework.

What is claimed is: 
     
       1. A method, comprising:
 receiving, by a device, information identifying a communication framework for a mass communication task; 
 determining, by the device, a success score for the communication framework using a mass communication model,
 wherein the success score represents a likelihood of a successful response in connection with using the communication framework for the mass communication task; 
 
 generating, by the device, a recommendation for the communication framework based on the success score and using the mass communication model; 
 altering, by the device, the communication framework to implement the recommendation and generate a modified communication framework,
 wherein altering the communication framework comprises:
 optimizing an assignment of message recipients to time slots based on determining a ratio of responses received to messages sent in other mass communication tasks that includes a particular message recipient, 
 selecting, using the mass communication model and based on end-user data, a first timing for messaging associated with a first message recipient,
 wherein selecting the first timing comprises: 
  determining a first set of time slot scores for the first message recipient, 
  wherein a time slot score, of the first set of time slot scores, represents a likelihood of response in a time slot of a set of time slots for messaging, 
  determining a second set of time slot scores for the second message recipient, and 
  optimizing an assignment of message recipients to time slots based on at least one of the first set of time slot scores, the second set of time slot scores, or a time slot availability criterion, and 
 
 selecting, using the mass communication model and based on the end-user data, a second timing that is different from the first timing for messaging associated with a second message recipient; and 
 
 
 performing, by the device, the mass communication task using the modified communication framework. 
 
     
     
       2. The method of  claim 1 , further comprising:
 receiving a mass communication data set identifying results of a set of mass communication tasks performed using a set of communication frameworks; 
 generating, using a machine learning technique and based on the mass communication data set, the mass communication model; and 
 storing, before receiving the information identifying the communication framework for the mass communication task, the mass communication model for subsequent use in evaluating the communication framework for the mass communication task. 
 
     
     
       3. The method of  claim 1 , wherein the recommendation relates to at least one of:
 a length of an initial message of the communication framework, 
 a relevance of the initial message of the communication framework, 
 a complexity of the communication framework, 
 a clarity of the communication framework, 
 a scheduling of messaging in accordance with the communication framework, or 
 a set of recipients of messages of the communication framework. 
 
     
     
       4. The method of  claim 1 , wherein the success score corresponds to a likelihood of achieving a threshold response rate to one or more messages of the communication framework. 
     
     
       5. The method of  claim 1 , wherein the mass communication model is trained using at least one of:
 a random forest classifier technique, 
 a multilayer perceptron technique, 
 a stochastic gradient descent technique, or 
 a neural network technique. 
 
     
     
       6. The method of  claim 1 , wherein the end-user data includes data identifying at least one of:
 a message recipient location, 
 a message recipient job role, 
 a message recipient job level, or 
 a message recipient response history. 
 
     
     
       7. A device, comprising:
 one or more memories; and 
 one or more processors communicatively coupled to the one or more memories, configured to:
 receive a mass communication data set identifying results of a set of mass communication tasks performed using a set of communication frameworks; 
 generate, using a machine learning technique and based on the mass communication data set, a mass communication model; 
 store the mass communication model for subsequent use in evaluating a communication framework for a mass communication task; 
 receive, after storing the mass communication model, information identifying the communication framework for the mass communication task; 
 determine a success score for the communication framework using the mass communication model,
 wherein the success score represents a likelihood of a successful response in connection with using the communication framework for the mass communication task; 
 
 generate a recommendation for the communication framework based on the success score and using the mass communication model; 
 alter the communication framework to implement the recommendation and generate a modified communication framework,
 wherein, the one or more processors, when altering the communication framework, are to:
 determine a ratio of responses received to messages sent in other mass communication tasks that includes a particular message recipient, 
 select, using the mass communication model and based on end-user data, a first timing for messaging associated with a first message recipient, 
  wherein the one or more processors, to select the first timing, are configured to: 
  determine a first set of time slot scores for the first message recipient, 
  wherein a time slot score, of the first set of time slot scores, represents a likelihood of response in a time slot of a set of time slots for messaging, 
  determine a second set of time slot scores for the second message recipient, and 
  optimize an assignment of message recipients to time slots based on at least one of the first set of time slot scores, the second set of time slot scores, or a time slot availability criterion, and 
  select, using the mass communication model and based on the end-user data, a second timing that is different from the first timing for messaging associated with a second message recipient; and 
 
 
 perform the mass communication task using the modified communication framework. 
 
 
     
     
       8. The device of  claim 7 , wherein the one or more processors are configured to:
 select, using the mass communication model and based on the end-user data, a first messaging channel for messaging associated with a third message recipient; and 
 select, using the mass communication model and based on the end-user data, a second messaging channel, that is different from the first messaging channel, for messaging associated with a fourth message recipient. 
 
     
     
       9. The device of  claim 8 , wherein the end-user data includes data identifying at least one of:
 a message recipient location, 
 a message recipient job role, 
 a message recipient job level, or 
 a message recipient response history. 
 
     
     
       10. The device of  claim 8 , wherein the one or more processors, when selecting the first messaging channel, are configured to:
 determine a first set of messaging channel scores for the first message recipient,
 wherein a messaging channel score, of the first set of messaging channel scores, represents a likelihood of response in a messaging channel of a set of messaging channels for messaging; 
 
 determine a second set of messaging channel scores for the second message recipient; and 
 optimize an assignment of message recipients to messaging channels based on at least one of the first set of messaging channel scores, the second set of messaging channel scores, or a messaging channel availability criterion. 
 
     
     
       11. The device of  claim 7 , wherein the one or more processors, when performing the mass communication task, are configured to:
 transmit a set of messages to a set of message recipients; 
 monitor for a set of responses to the set of messages; and 
 selectively provide follow-up messages to one or more message recipients of the set of message recipients, based on a result of monitoring for the set of responses to the set of messages. 
 
     
     
       12. The device of  claim 7 , wherein the one or more processors, when altering the communication framework, are configured to:
 determine, based on the mass communication model, at least one message recipient, of a set of message recipients to whom a message is to be sent, to omit from the mass communication task. 
 
     
     
       13. A non-transitory computer-readable medium storing instructions, the instructions comprising:
 one or more instructions that, when executed by one or more processors, cause the one or more processors to:
 receive information identifying a communication framework for a mass communication task; 
 determine a success score for the communication framework using a mass communication model,
 wherein the success score represents a likelihood of a successful response in connection with using the communication framework for the mass communication task; 
 
 generate a recommendation for the communication framework based on the success score and using the mass communication model,
 wherein the recommendation relates to message recipient-level differentiation such that a first message recipient is associated with a different time slot, messaging channel, or message content relative to a second message recipient; 
 
 alter the communication framework to implement the recommendation and generate a modified communication framework,
 wherein the one or more instructions, that cause the one or more processors to alter the communication framework, cause the one or more instructions to:
 optimize an assignment of message recipients to time slots based on determining a ratio of responses received to messages sent in other mass communication tasks that includes a particular message recipient, 
 select, using the mass communication model and based on end-user data, a first timing for messaging associated with the first message recipient, 
  wherein the one or more instructions, that cause the one or more processors to select the first timing, cause the one or more processors to: 
  determine a first set of time slot scores for the first message recipient, 
  wherein a time slot score, of the first set of time slot scores, represents a likelihood of response in a time slot of a set of time slots for messaging, 
  determine a second set of time slot scores for the second message recipient, and 
  optimize an assignment of message recipients to time slots based on at least one of the first set of time slot scores, the second set of time slot scores, or a time slot availability criterion, and 
 select, using the mass communication model and based on the end-user data, a second timing that is different from the first timing for messaging associated with the second message recipient; and 
 
 
 perform the mass communication task using the modified communication framework. 
 
 
     
     
       14. The non-transitory computer-readable medium of  claim 13 , wherein the one or more instructions, that cause the one or more processors to alter the communication framework, cause the one or more processors to:
 determine, based on the mass communication model, at least one message recipient, of a set of message recipients to whom a message is to be sent, to omit from the mass communication task. 
 
     
     
       15. The non-transitory computer-readable medium of  claim 13 , wherein the one or more instructions, when executed by the one or more processors, further cause the one or more processors to:
 determine a set of message recipient responses to the mass communication task; 
 determine, based on the set of message recipient responses, a set of engagement scores for a set of message recipients associated with the mass communication task; 
 generate an engagement recommendation for message recipient retention based on the set of engagement scores; and 
 automatically perform a response action to implement the engagement recommendation. 
 
     
     
       16. The non-transitory computer-readable medium of  claim 15 , wherein the set of engagement scores is determined based on at least one of:
 a rate of response to messaging of the mass communication task, 
 an engagement with the mass communication task, 
 a rate of compliance with directions of the mass communication task, 
 a content of the set of message recipient responses, or 
 a timeliness of the set of message recipient responses. 
 
     
     
       17. The non-transitory computer-readable medium of  claim 15 , wherein an engagement score, of the set of engagement scores, corresponds to a predicted rate of attrition for a message recipient of the set of message recipients. 
     
     
       18. The non-transitory computer-readable medium of  claim 13 , wherein the one or more instructions, that cause the one or more processors to receive the information identifying the communication framework, cause the one or more processors to:
 receive information identifying a first part of the communication framework; 
 generate, based on the mass communication model and the first part of the communication framework, a communication framework recommendation for a second part of the communication framework; and 
 automatically complete the communication framework based on the communication framework recommendation. 
 
     
     
       19. The method of  claim 1 , further comprising:
 selecting, using the mass communication model and based on the end-user data, a first messaging channel for messaging associated with the first message recipient; and 
 selecting, using the mass communication model and based on the end-user data, a second messaging channel, that is different from the first messaging channel, for messaging associated with the second message recipient. 
 
     
     
       20. The method of  claim 1 , wherein altering the communication framework, comprises:
 determining, based on the mass communication model, at least one message recipient, of a set of message recipients to whom a message is to be sent, to omit from the mass communication task.